Herschel–ATLAS: First data release of the Science ... · Herschel–ATLAS: First data release of...

Mon. Not. R. Astron. Soc. 000, 1–13 (2002) Printed 29 October 2018 (MN LATEX style file v2.2)

Herschel–ATLAS: First data release of the Science DemonstrationPhase source catalogues

E.E. Rigby1?, S.J. Maddox1, L. Dunne1, M. Negrello2, D.J.B. Smith1, J. Gonzalez-Nuevo3, D. Herranz4, M. Lopez-Caniego4, R. Auld5, S. Buttiglione6, M.Baes11, A.Cava7, A. Cooray8, D. L. Clements9, A. Dariush5, G. De Zotti3,6, S. Dye5, S. Eales5, D.Frayer10, J. Fritz11, R. Hopwood2, E. Ibar12, R.J. Ivison12,13, M. Jarvis14, P. Panuzzo15,E. Pascale5, M. Pohlen5, G. Rodighiero6, S. Serjeant2, P. Temi16, M. A. Thompson14

1School of Physics and Astronomy, University of Nottingham, University Park, Nottingham NG7 2RD, UK2Department of Physics and Astronomy, The Open University, Walton Hall, MK7 6AA Milton Keynes, UK3SISSA, Via Bonomea 265, I-34136 Trieste, Italy4Instituto de Fısica de Cantabria (CSIC-UC), Avda. los Castros s/n, 39005 Santander, Spain5School of Physics and Astronomy, Cardiff University, The Parade, Cardiff, CF24 3AA, UK6INAF Osservatorio Astronomico di Padova, Vicolo Osservatorio 5, I-35122 Padova, Italy7Departamento de Astrofısica, Facultad de CC. Fısicas, Universidad Complutense de Madrid, E-28040 Madrid, Spain8Department of Physics and Astronomy, University of California, Irvine, CA 92697, USA9Astrophysics Group, Imperial College, Prince Consort Road, London SW7 2AZ, UK10Infrared Processing and Analysis Center; California Institute of Technology 100-22, Pasadena, CA 91125, USA11Sterrenkundig Observatorium, Universiteit Gent, Krijgslaan 281 S9, B-9000 Gent, Belgium12UK Astronomy Technology Centre, Royal Observatory Edinburgh, Edinburgh, EH9 3HJ, UK13Institute for Astronomy, University of Edinburgh, Royal Observatory, Edinburgh, EH9 3HJ, UK14Centre for Astrophysics, Science & Technology Research Institute, University of Hertfordshire, Hatfield, Herts, AL10 9AB, UK15Centre CEA de Saclay (Essonne), Gif-sur-Yvette, 921191 cedex, France16Astrophysics Branch, NASA Ames Research Center, Mail Stop 245-6, Moffett Field, CA 94035, USA

ABSTRACTThe Herschel–ATLAS is a survey of 550 square degrees with the Herschel Space Observatoryin five far–infrared and submillimetre bands. The first data for the survey, observations of afield 4 × 4 deg2 in size, were taken during the Science Demonstration Phase, and reach a 5σnoise level of 33.5 mJy/beam at 250µm . This paper describes the source extraction meth-ods used to create the corresponding Science Demonstration Phase catalogue, which contains6876 sources, selected at 250µm , within ∼14 sq. degrees. SPIRE sources are extracted usinga new method specifically developed for Herschel data; PACS counterparts of these sourcesare identified using circular apertures placed at the SPIRE positions. Aperture flux densitiesare measured for sources identified as extended after matching to optical wavelengths. Thereliability of this catalogue is also discussed, using full simulated maps at the three SPIREbands. These show that a significant number of sources at 350 and 500µm have undergoneflux density enhancements of up to a factor of ∼2, due mainly to source confusion. Correctionfactors are determined for these effects. The SDP dataset and corresponding catalogue will beavailable from http://www.h-atlas.org/.

Key words:

1 INTRODUCTION

The Herschel Astrophysical Terahertz Large Area Survey (H-ATLAS) survey is the largest, in time and area, of the extragalacticOpen Time Key Projects to be carried out with the European Space

? E-mail: [email protected]; [email protected]

Agency (ESA) Herschel Space Observatory (Pilbratt et al. 2010)1.When complete it will cover ∼550 square degrees of the sky, infive far–infrared and submillimetre bands (100, 160, 250, 350 and

1 Herschel is an ESA space observatory with science instruments providedby European-led Principal Investigator consortia and with important partic-ipation from NASA.

c© 2002 RAS

arX

iv:1

010.

5787

v2 [

astr

o-ph

.IM

] 7

Apr

201

1

2 E.E. Rigby

500µm ), to a 5σ depth of 33 mJy/beam at 250µm . The predictednumber of sources is ∼200,000; of these ∼40,000 are expected tolie within z < 0.3. A full description of the survey can be found inEales et al. (2010).

This paper presents the 250µm selected source cataloguecreated from the initial H-ATLAS Science Demonstration Phase(SDP) observations. Eight papers based on this catalogue have al-ready been published in the A&A Herschel Special Issue rangingfrom the identification of blazars (Gonzalez-Nuevo et al. 2010) anddebris disks (Thompson et al. 2010) in the SDP field, to determina-tions of the colours (Amblard et al. 2010), source counts (Clementset al. 2010), clustering (Maddox et al. 2010) and 250µm luminosityfunction evolution (Dye et al. 2010) of the submillimetre popula-tion, as well as the star formation history of quasar host galaxies(Serjeant et al. 2010) and the dust energy balance of a nearby spiralgalaxy (Baes et al. 2010).

The layout of the paper is as follows: Section 2 describes theSDP observations; Section 3 describes the source extraction proce-dure for the five bands; finally, Section 4 outlines the simulationsused to quantify the reliability of the catalogue. For more details ofthe SDP data see Pascale et al. (2010) and Ibar et al. (2010a) forthe SPIRE and PACS data reduction respectively, and Smith et al.(2011) for the multiwavelength catalogue matching.

2 HERSCHEL OBSERVATIONS

The SDP observations for the H–ATLAS survey cover an area of∼4◦×4◦, centred at α=09h05m30.0s, δ =00◦30′ 00.0′′ (J2000).This field lies within one of the regions of the GAMA (Galaxy andMass Assembly) survey (Driver et al. 2009) so optical spectra,along with additional multiwavelength data, are available for themajority of the low–redshift sources.

The observations were taken in parallel–mode, which uses thePhotodetector Array Camera and Spectrometer (PACS; Poglitsch etal. 2010) and Spectral and Photometric Imaging REciever (SPIRE;Griffin et al. 2010) instruments simultaneously; two orthogonalscans were used to mitigate the effects of 1/f noise. The time–linedata were reduced using HIPE (Ott et al. 2010). SPIRE 250, 350,and 500µm maps were produced using a naıve mapping technique,after removing any instrumental temperature variations (Pascale etal. 2010), and incorporating the appropriate flux calibration factors.Noise maps were generated by using the two cross–scan measure-ments to estimate the noise per detector pass, and then for eachpixel the noise is scaled by the square root of the number of detec-tor passes. The SPIRE point spread function (PSF) for each bandwas determined from Gaussian fits to observations of Neptune, theprimary calibrator for the instrument. Maps from the PACS 100 and160µm data were produced using the PhotProject task withinHIPE (Ibar et al. 2010a). A false colour combined image of a partof the three SPIRE maps is shown in Figure 1. The measured beamfull–width–half–maxima (FWHMs) are approximately 9′′ , 13′′ ,18′′ , 25′′ and 35′′ for the 100, 160, 250, 350 and 500µm bands re-spectively (Ibar et al. 2010a; Pascale et al. 2010). The map pixelsare 2.5′′ , 5′′ , 5′′ , 10′′ and 10′′ in size for the same five bands.

The noise levels measured by Pascale et al. (2010) for the250µm and 500µm SPIRE bands are in good agreement with thosepredicted using the Herschel Space Observatory Planning Tool(HSpot2); for the 350µm band they are considerably better. The

2 HIPE and HSpot are joint developments by the Herschel Science Ground

(a) Original combined map

(b) After background subtraction

Figure 1. False–colour images of a 1.5 sq. degree region of the SDP fieldshowing the three SPIRE bands combined. Image (a) is before background–subtraction and shows clear contamination by galactic cirrus; image (b)shows the reduction in contamination after subtracting the background.

corresponding PACS noise levels determined by Ibar et al. (2010a)are currently higher than predicted (26 mJy and 24 mJy, comparedwith 13.4 mJy and 18.9 mJy for 100µm and 160µm respectively),but this may improve in future with better map–making tech-niques. The flux calibration uncertainties are 15% for the threeSPIRE bands (Pascale et al. 2010) and 10 and 20% for the PACS100µm and 160µm bands respectively (Ibar et al. 2010a).

3 SOURCE EXTRACTION

The ultimate aim for the source identification of the H-ATLASdata is to use a multiband method to perform extraction across thefive wavebands simultaneously, thus utilising all the available dataas well as easily obtaining complete flux density information foreach detected galaxy, without having to match catalogues betweenbands. However, the short timescale for the reduction of these SDPobservations, combined with the higher than expected PACS noise

Segment Consortium, consisting of ESA, the NASA Herschel Science Cen-ter, and the HIFI, PACS and SPIRE consortia

c© 2002 RAS, MNRAS 000, 1–13

The H-ATLAS SDP catalogue 3

Ext S/N = 14.11

Noiseless map

trueext

Ext S/N = 14.11

Data map

trueext

Ext S/N = 14.11

PSF Filtered map

trueext

(a) Source with a close companion

Ext S/N = 32.43

Noiseless map

trueext

Ext S/N = 32.43

Data map

trueext

Ext S/N = 32.43

PSF Filtered map

trueext

(b) Isolated source

Figure 2. The input (true) and extracted position for two point sources in the 250µm simulated maps before the addition of Gaussian noise (noiseless), afterthe noise has been added (noisy), and after further convolving with the 250µm point spread function (PSF) to create the final realistic sky (PSF–filtered) (seeSection 4 for full details), to illustrate how the position, and therefore flux density, of an extracted source found by MADX can be influenced by the presenceof a close companion.

250 µm

10 100 1000Aperture S250 (mJy)

10

100

1000

MA

DX

S25

0 (m

Jy)

All sourcesExtended sources

350 µm

10 100 1000Aperture S350 (mJy)

10

100

1000

MA

DX

S35

0 (m

Jy)


500 µm

10 100Aperture S500 (mJy)

10

100

MA

DX

S50

0 (m

Jy)


Figure 3. A comparison between the MADX and aperture measured fluxes for the sources with a possible optical identification in the matched catalogue ofSmith et al. (2011). Source identified as extended are highlighted in bold.

levels, means that this was only possible for the three SPIRE bands.As a result, the source extraction for the PACS and SPIRE maps isdiscussed separately in this Section.

The full H-ATLAS SDP catalogue described here will beavailable at http://www.h-atlas.org/.

3.1 The SPIRE catalogue

Sources are identified in the SPIRE 250, 350 and 500µm maps us-ing the Multi–band Algorithm for source eXtraction (MADX, Mad-dox et al. 2011), which is being developed for the H–ATLAS sur-vey. Several methods for generating the final SPIRE catalogue withMADX were investigated and these are described below.

The first step in the MADX source extraction is to subtract

a local background, estimated from the peak of the histogram ofpixel values in 30 × 30 pixel blocks (chosen to allow the mapto be easily divided up into independent sub–regions). This cor-responds to 2.5′× 2.5′ for the 250µm map, and 5′× 5′ for the 350and 500µm maps. The background (in mJy/beam) at each pixel wasthen estimated using a bi-cubic interpolation between the coarsegrid of backgrounds, and subtracted from the data. Figure 1 il-lustrates the reduction in background contamination (mainly aris-ing from galactic cirrus, which dominates over the confusion noisefrom unresolved sources) obtained using this method.

The background subtracted maps were then filtered by theestimated PSF, including an inverse variance weighting, wherethe noise for each map pixel was estimated from the noise map(matched filtering, e.g. Turin 1960; Serjeant et al. 2003). The back-

c© 2002 RAS, MNRAS 000, 1–13

4 E.E. Rigby

ground removal has a negligible effect on the PSF because the his-togram peak is insensitive to resolved sources in the backgroundaperture; this will be discussed further in Maddox et al. (2011).We also create a ‘filtered noise’ map which represents the noise ona pixel in the PSF filtered map. This is lower than the raw noisemap because the noise in the SPIRE pixels is uncorrelated, and sofiltering by the PSF reduces the noise by approximately the squareroot of the number of pixels per beam.

The maps from the 350 and 500µm bands are interpolatedonto the 250µm pixels. Then all three maps are combined withweights set by the local inverse variance, and the prior expec-tation of the spectral energy distribution (SED) of the galaxies.We used two SED priors: a flat-spectrum prior (assumed to beflat in fν ), where equal weight is given to each band; and also250µm weighting, where only the 250µm band was included.

Local, > 2.5σ, peaks are identified in the combined PSF fil-tered map as potential sources, and sorted in order of decreasingsignificance level. A Gaussian is fitted to each peak in turn to pro-vide an estimate of the position at the sub–pixel level; this can beinfluenced by the presence of a neighbouring source, as illustratedin Figure 2, but the effect is minimal. The flux in each band is thenestimated using a bi–cubic interpolation to the position given bythe combined map. The scaled PSF is then subtracted from the mapbefore going on to the next source in the sequence. This ensuresthat flux from the wings of bright sources does not contaminatenearby fainter sources. This sorting and PSF subtraction reducesthe effect of confusion, but in future releases we plan to implementmulti-source fitting to blended sources.

To produce a catalogue of reliable sources, a source is onlyincluded if it is detected at a significance of at least 5σ in one of theSPIRE bands. The total number of sources in the SPIRE catalogueis 6876.

For our current data we chose to use the 250µm only priorfor all our catalogues, which means that sources are identified at250µm only. At the depth of the filtered maps source confusion isa significant problem, and the higher resolution of the 250µm mapsoutweighed the signal–to–noise gain from including the otherbands (see Section 4.1 and Figure 8c). This may introduce a bias inthe catalogue against red, potentially high–redshift, sources that arebright at 500µm , but weak in the other bands. However, compar-ing catalogues made with both the 250µm and flat–spectrum pri-ors showed that the number of missed sources is low: 2974 > 5σ350µm sources and 348 > 5σ 500µm sources are detected withthe flat prior, compared with 2758 and 307 sources detected us-ing the 250µm prior (i.e. 7% and 12% of sources are missed at350µm and 500µm respectively). It should also be noted that fora high–redshift source to be missed it would need a 500µm to250µm flux ratio of > 2.7 (i.e. it has to be < 2.5σ at 250µm tobe excluded from the catalogue). Assuming typical SED templates(e.g. M82 and Arp220), this means that this should only occur forsources which lie at redshifts > 4.6. We aim to revisit this issue infuture data–releases.

Since MADX uses a bicubic interpolation to estimate the peakflux in the PSF filtered map, it partially avoids the peak suppressioncaused by pixelating the time-line data, as discussed by Pascale etal. Nevertheless the peak fluxes are systematically underestimated,and so pixelization correction factors were calculated by pixelatingthe PSF at a large number of random sub-pixel positions. The meancorrection factors were found to be 1.05, 1.11 and 1.04 in the 250,350 and 500µm bands respectively, and they have been included inthe released SDP catalogue.

In calculating the σ for each source, we use the filtered noise

100 µm

1 10 100 1000S / mJy

104

105

dN

/dS

* S

2.5 /

de

g-2 m

Jy

1.5

PEP: Lockman HolePEP: GOODS NPEP: COSMOSH-ATLAS: GAMA9

160 µm

1 10 100 1000S / mJy

104

105

106

dN

/dS

* S

2.5 /

de

g-2 m

Jy

1.5

PEP: Lockman HolePEP: GOODS NPEP: COSMOSH-ATLAS: GAMA9

Figure 4. The differential source counts from the PACS section of the SDPcatalogue compared to the initial results from the three fields covered by thePEP survey (Berta et al. 2010).

1IRAS S100µm / Jy

1

H-A

TLA

S

S10

0µm /

Jy

Figure 5. A comparison between the 100µm flux densities from PACS andIRAS

map and add the confusion noise to this in quadrature. The average1σ instrumental noise values are 4.1, 4.0 and 5.7 mJy/beam respec-tively, with 5% uncertainty, in the 250, 350 and 500µm bands, de-termined from the filtered maps (Pascale et al. 2010). We estimatedthe confusion noise from the difference between the variance of themaps and the expected variance due to instrumental noise (assum-ing that confusion is dominating the excess noise), and find that the1σ confusion noise is 5.3, 6.4 and 6.7 mJy/beam at 250, 350 and500µm , with an uncertainty of 7%; these values are in good agree-ment with those found by Nguyen et al. (2010) using data from theHerschel Multi–tiered Extragalactic Survey (HerMES). The result-ing average 5σ limits are therefore 33.5, 37.7 and 44.0 mJy/beam.

c© 2002 RAS, MNRAS 000, 1–13


3.1.1 Extended sources

The flux density extracted by MADX will underestimate thetrue value for sources that are larger than the SPIRE beams,which have FWHM of 18.1′′ , 24.8′′ and 35.2′′ for 250, 350 and500µm respectively. This occurs because the peak value taken byMADX only accurately represents the true flux density of a sourceif it is point–like. These extended sources can be identified if theyalso have a reliable optical match and therefore a correspondingoptical size, ropt (equivalent to the 25 mag arcsec−2 isophote), inthe SDSS or GAMA catalogues (see Smith et al. (2011) for full de-tails of the matching procedure and the determination of the matchreliability, Rj). The size of the aperture used is listed in the cata-logue, and the most appropriate flux density, either point source oraperture measurement (when this is larger), is given for each sourcein the SPIRE ‘BEST flux’ columns. It should be noted that, apartfrom two exceptions, this is necessary at 250 and 350µm only, asthe large 500µm beam size means that the flux discrepancy is neg-ligible for that map.

An ‘extended source’, in a particular map, is defined here asone with ropt > 0.5×FWHM, and to ensure only true matchesare used, it must also have a match–reliability, Rj , greater than0.8. In total, the MADX ‘BEST’ flux columns for 167 sources at250µm and 53 sources at 350µm were updated with aperture pho-tometry values.

The aperture radius, ar , in a particular band is set by summingthe optical size in quadrature with the FWHM of that band:

ar =√

FWHM2 + r2opt. (1)

The exceptions to this were the apertures used for sources H–ATLAS J091448.7-003533 (a merger, where the given ar isinsufficient to include the second component) and H–ATLASJ090402.9+005436, which visual inspection showed was clearlyextended. In these cases the aperture sizes used are chosen to matchthe extent of the sub–mm emission, and fluxes are replaced in the500µm band as well.

The apertures are placed on the MADX, Jy/beam, backgroundsubtracted maps, at the catalogue position for each source; the mea-sured values are converted to the correct flux scale by dividing bythe area of the beam derived by Pascale et al. (2010) for each map(13.9, 6.6 or 14.2 pixels for 250, 350 and 500µm respectively). Thecorresponding 1σ error is given by √vap, where vap is the sumof the variances within apertures placed in the same positions onthe relevant variance maps. Confusion noise estimates were againadded in quadrature to these uncertainties; these were scaled ac-cording to the area of each individual aperture.

Figure 3 compares the MADX and aperture measured fluxesfor all catalogue sources with a possible optical identification. Itshows that the majority of objects are point–like, for which theagreement between the two sets of fluxes is good. The sources iden-tified as extended are highlighted in bold, and it is clear that MADXunderestimates these at 250 and 350µm if they are brighter than∼100 mJy.

3.2 The PACS catalogue

The higher noise levels in the PACS maps, along with the shapeof the source SEDs, mean that all the PACS extragalactic sourcesshould be clearly detected in the SPIRE catalogue. Sources in thePACS data are therefore identified by placing circular apertures atthe SPIRE 250µm positions in the 100µm and 160µm maps, aftercorrecting the PACS astrometry to match that of the 250µm map

(using the sources present in both the SPIRE and PACS maps).There are two steps to this source detection process: first a ‘pointsource’ measurement is obtained for all SPIRE positions usingapertures with radii of 10′′ (100µm ) or 15′′ (160µm ); next addi-tional aperture fluxes are found for positions where a PACS sourcewould satisfy the extended source criteria discussed in Section3.1.1. Aperture radii in this case are calculated using Equation 1, as-suming FWHM of 8.7′′ and 13.1′′ for 100 and 160µm respectively.These FWHM values are calculated using rough modelling of theVesta asteroid as the full PACS PSFs are asymmetric (see Ibar et al.2010a, for a full discussion).

The aggressive filtering used for these maps means that thelarge scale structure in the cirrus has already been removed,but some noise stripes remain. These are removed globally at160µm by subtracting a background determined within 10×10pixel blocks. However, at 100µm this global approach was found tointroduce negative holes around bright sources so the backgroundvalue is determined for each source individually using a local an-nulus with a width of 0.5 times the aperture radius.

Unlike SPIRE, the PACS maps have units of Jy/pixel so nobeam conversion is needed. However, the fluxes are divided by1.09 (100µm ) or 1.29 (160µm ) as recommended by the PACSInstrument Control Centre3. These scaling factors are now incor-porated into the data–reduction pipeline and have been applied tothe public release of the PACS SDP maps, along with the astrome-try correction needed to match that of the SPIRE 250µm map (thiscorrection is ∼1′′ in both PACS bands). The fluxes are also aper-ture corrected, using a correction determined from observations ofa bright point–like source. The 1σ errors are found using aperturesrandomly placed in the maps; note that these errors scale with aper-ture size. The low confusion noise compared to SPIRE, plus the fastscan speed used in these observations, means that the integrationtime used in H-ATLAS is insufficient to provide confusion limitedimages with PACS. Full details of these observations can be foundin Ibar et al. (2010a).

The most appropriate flux density measurements, either pointor extended (where this is larger), are given in the ‘BEST’ PACScolumns in the SDP catalogue, along with the corresponding aper-ture radii, for sources with S/N > 5. As a result 151 and 304sources satisfy this condition at 100 and 160µm respectively. The5σ point source limits in the PACS catalogue are 132 mJy and121 mJy at 100 and 160µm . It should be noted that the fluxdensities extracted from the PACS maps are only at 100µm and160µm under the assumption of a constant energy spectrum,though the colour corrections for sources with a different SED aresmall (Poglitsch et al. 2010).

The PACS time-line data have been high-pass filtered by sub-tracting a boxcar median over 3.4 arcmin (at 100µm ) and 2.5′ at160µm (Ibar et al. 2010a). The filtering will lead to the underes-timation of flux for sources extended on scales comparable to thefilter length. The exact flux loss for a particular source will de-pend on the size of the source along the scan directions, and willalso depend on whether the peak surface brightness is above the4σ threshold used in the second level filtering. A simple simula-tion of a circular exponential disc shows that the filtering removes∼ 50% of the source flux if the diameter of the disk is equal to thefilter length. If the diameter is half of the filter length, then only5% of the flux is removed. This suggests that sources with a diame-ter less than 1′ should by relatively unaffected by the filtering. Flux

3 see the scan mode release note, PICCMETN0.35

c© 2002 RAS, MNRAS 000, 1–13

6 E.E. Rigby

0.1 1.0Flux density (Jy)

0.01

0.10

1.00

10.00

100.00

1000.00

N(>

S)

deg-2

True catalogueExtracted catalogueReal SD catalogue

500 µm350 µm250 µm

(a) Extended source simulations

0.1 1.0Flux density (Jy)

0.01

0.10

1.00

10.00

100.00

1000.00

N(>

S)

deg-2

True catalogueExtracted catalogueReal SD catalogue

500 µm350 µm250 µm

(b) Point source only simulations

Figure 6. The integrated source counts from the combined set of 500 input(true) and extracted simulated catalogues, along with those calculated usingthe SDP catalogue for both versions of the simulations.

250 µm

-20 -10 0 10 20

True RA - Extracted RA

-20

-10

0

10

20

Tru

e D

EC

- E

xtr

acte

d D

EC

Figure 7. The positional offsets between the matched sources in the simu-lated extracted and input (true) full 250µm catalogues. The results for thetwo versions are very similar, so only the PSS points are shown here.

measurements for sources larger than this should be treated withcaution.

Figure 4 compares the differential source counts calculatedfrom the PACS SDP catalogues to those determined from the initialdata of the complementary PACS Evolutionary Probe (PEP) survey(Berta et al. 2010), which is deeper than H–ATLAS but covers asmaller area. The good agreement between the two sets of countssupports the initial assumption that all bright PACS sources shouldalready be present in the SPIRE catalogue. However, there are in-sufficient sources in the SDP data to properly constrain the brightnumber counts tail. A full analysis of the PACS counts will be pre-sented in (Ibar et al. 2010b).

For the sources detected in the PACS 100µm map an addi-tional comparison can be made to this wavelength in the Impe-rial IRAS–FSC Redshift Catalogue of Wang & Rowan-Robinson(2009), which combines the original IRAS Faint Source Catalogueflux density values with improved optical and radio identificationsand redshifts. There are 34 IRAS sources within the PACS regionof the H-ATLAS SDP field; 19 of these have a reliable IRAS fluxmeasurement and these are in good agreement with the SDP cata-logue, with a mean offset consistent with zero, as shown in Figure5.

4 ASSESSING THE CATALOGUE RELIABILITY

4.1 Simulation creation

It is not enough to identify sources in the H-ATLAS SDP maps; therobustness of the catalogue must also be determined. This is doneusing realistic simulations of the observations, with the same noiseproperties as the processed maps, and a realistic cirrus background,based on IRAS measurements (Schlegel et al. 1998). However, onlythe three SPIRE bands are considered in this initial analysis, asthe PACS SDP catalogue is currently treated as an extension to theSPIRE data

The simulated maps are randomly populated with sources gen-erated using the models of Negrello et al. (2007), which predict thenumber counts of both the spheroidal and protospheroidal galaxypopulations separately; for the simulations, these predictions arecombined together to give the expected total counts, and hencethe corresponding set of source flux densities, for each band. Al-though Maddox et al. (2010) detected, in SDP data, strong cluster-ing for 350µm and 500µm–selected samples, fluctuations due tofaint sources at the SPIRE resolution are Poisson dominated, espe-cially at 250µm (e.g. Negrello et al. 2004; Viero et al. 2009). Thissuggests that, for the present purposes, using unclustered randompositions is a sufficiently good approximation. The flux densitiesof all the sources in the models are reduced by 26% at 250µm and15% at 350µm to improve the agreement with the observed (i.e.uncorrected) source counts in the SDP catalogue (Clements et al.2010); the results of this alteration are shown in Figure 6. The fi-nal flux density ranges are 0.11 mJy – 1.65 Jy at 250µm , 0.24mJy – 0.83 Jy at 350µm and 0.45 mJy – 0.59 Jy at 500µm for thesimulated sources; this ensures that the simulated maps contain arealistic background of faint sources which can contribute to theconfusion noise.

The simulations are constructed by first adding the flux of eachsource in each band to the relevant position in a 1 arcsecond grid.Two versions of the simulations are created. In the first the sim-ulated sources are all one pixel in size (point–source–simulations:PSS), whereas in the second the sources are assigned a scale–length

c© 2002 RAS, MNRAS 000, 1–13


10 100Extracted S/N

0.0

0.5

1.0

1.5

2.0

2.5

Pos

ition

al e

rror

/ ar

csec

σ(∆(Dec))σ(∆(RA))


10 100Extracted S/N

0.0

0.5

1.0

1.5

2.0

2.5

Pos

ition

al e

rror

/ ar

csec


Smith et al predictionSmith et al measurement

Positional error plotted against True S/N

(b) Point source simulations

10 100Extracted S/N

0.0

0.5

1.0

1.5

2.0

2.5

3.0

Pos

ition

al e

rror

/ ar

csec


250 µm priorFlat spectrum prior

(c) Point source simulations comparing different priors

Figure 8. The positional errors for the two different versions of the simulations, alongside a comparison of the two different source extraction position priors aspreviously discussed in Section 3.1. Also shown in 8b are the positional errors plotted against the S/N in the input (true) catalogue, along with those determinedby Smith et al. (2011) for the SDP data at 5, 7.5 and 10σ.

based on their catalogue redshift (extended–source–simulations:ESS). The scale–length is constant in physical units, and then con-verted to an angular scale using standard cosmology. The ESS willobviously be a better representation of the real data, but, as Sec-tion 3.1.1 shows, MADX underestimates the flux densities of ob-jects with sizes larger than the FWHM, so the PSS simulations pro-vide a useful comparison. It should be noted that the flux densitiesand positions of the input sources will be the same in both cases.The next step is to convolve the 1 arcsecond map by the appropri-ate Herschel PSF, also sampled on a 1 arcsecond grid, to give amap of flux per beam covering the full area of the SDP data. Then,the 1 arcsecond pixels are block averaged to give 5 arcsecond pix-els for the 250µm maps, and 10 arcsecond pixels for the 350 and500µm maps.

A background representing emission from Galactic cirrus isthen added to the each map. The background value is estimatedfrom the Schlegel et al. (1998) map of 100µm dust emission andtemperature by assuming a modified black-body spectrum withβ = 2.0, and scaling to the appropriate wavelength. The resolu-tion of this IRAS map is lower than that in the SDP data, whichmeans that small scale structure in the cirrus is not present in thesimulations. Since the cirrus is highly structured, it is non–trivial togenerate realistic structure on smaller scales, so as a simple approx-imation, the low resolution maps were used, though it should benoted that the true cirrus background will include more small scalefeatures. It is clear that the real cirrus structure in the SDP data ishighly non–Gaussian, so simply extrapolating the power spectrumto smaller scales does not significantly improve the model back-ground.

Finally instrumental noise is added to each pixel as a Gaussiandeviate, scaled using the real coverage maps so that the local rms isthe same as in the real data.

Sources are then extracted with MADX from both versions ofthe simulations, following the procedure described in Section 3.1.For the ESS maps, the flux densities in the three bands are againreplaced with aperture–measured values for the extended sources.The ‘optical sizes’ (needed to determine ar using Equation 1) inthis case are taken as three times the scale–size taken from the in-put catalogues; this corresponds to a B–band isophotal limit of∼25mag arcsec−2 (Zhong et al. 2008). Finally, the MADX catalogue iscut to only include sources which are detected at the 5σ level inany of the available bands. This process is repeated 500 times, eachtime using a different realisation of the input model counts, to en-

sure sufficient numbers of bright sources are present at the longerwavelengths. The average number of extracted sources which arealso >5σ in any band is 5881 and 5772 for PSS and ESS respec-tively, which is lower than the 6876 sources present in the real SDPdata; as Figure 6 illustrates, this is because the simulated sourcecounts do not exactly reproduce the real SDP ones. Additionally,more sources are found for the PSS version because of the fluxunderestimation of extended sources which means that the faintestobjects fall below the catalogue cut. In the remainder of this discus-sion, these MADX catalogues will be referred to as the ‘extractedcatalogues’, and the simulated input source lists as the ‘simulatedinput catalogues’.

For each of the three bands in turn, starting with the bright-est, sources in the extracted catalogue are matched to the simu-lated input source that makes the largest contribution, determinedby weighting with the filtered beam, at that extracted position. Amatch radius of 3 pixels (approximately equal to the FWHM ineach band) is also imposed to ensure that a match is not made toan unfeasibly distant source. Since the typical positional error for a> 5σ 250µm source is 2.5′′ or less, this match radius will ensurethat almost no real matches are rejected, whilst the weighting willavoid spurious matches. Once matched, a simulated input source isremoved from consideration to avoid double–matches. Consideringeach band separately will allow an extracted source to have threedifferent simulated input counterparts, depending on where the ma-jority of its flux density comes from at 250, 350 and 500µm . Thisensures that the effects of source blending in the data can be prop-erly investigated, though it should be noted that the results are verysimilar if the counterparts are found at the highest resolution, short-est wavelength only. Full simulated input, extracted and matchedcatalogues for each band are then made by combining the resultsfrom the 500 individual sets of simulations together.

The positional offsets and corresponding errors are shown inFigures 7 and 8. They demonstrate that there is no significant offsetbetween the extracted and matched catalogues. The positional er-rors for 5σ sources are ∼2.4′′ at 250µm in both versions, whichagrees with the value of 2.40 ± 0.11′′ found for the real SDPdata by Smith et al. (2011). The errors also approximately scaleas 1/(S/N) in the 250µm band, as predicted by e.g. Ivison et al.(2007). However, at low S/N there is an enhancement over the pre-dicted values, as illustrated for the PSS results in Figure 8b. This isa result of Eddington bias causing more faint sources errors to scat-ter up than vice–versa; if the positional errors are plotted against the

c© 2002 RAS, MNRAS 000, 1–13

8 E.E. Rigby

250 µm

10Extracted S/N

1Sex

trac

ted/

Str

ue

6050.4

53 sigma clipped mean

median

350 µm

10Extracted S/N

1Sex

trac

ted/

Str

ue

6050.4


median

500 µm

10Extracted S/N

1Sex

trac

ted/

Str

ue

6050.4


median


250 µm

10Extracted S/N

1Sex

trac

ted/

Str

ue

6050.4


median

350 µm

10Extracted S/N

1Sex

trac

ted/

Str

ue

6050.4


median

500 µm

10Extracted S/N

1Sex

trac

ted/

Str

ue

6050.4


median


Figure 9. The ratio of flux densities for the matched sources in the simulated input (true) and extracted catalogues as a function of extracted signal to noise(S/N) for the three bands from the ESS and PSS maps. Also shown are the median and 3σ clipped mean values, calculated in bins of 0.05 in log(S/N).

250 µm

10Extracted S/N

1

Sex

trac

ted/

Sno

isel

ess

6050.5


median

350 µm

10Extracted S/N

1

Sex

trac

ted/

Sno

isel

ess

6050.5


median

500 µm

10Extracted S/N

1

Sex

trac

ted/

Sno

isel

ess

6050.5


median

Figure 10. The ratio of flux densities for the matched sources in the noiseless MADX (Snoiseless) and extracted catalogues as a function of extracted signalto noise for the three bands (including point sources only). Also shown are the median and 3σ clipped mean values, calculated in bins of 0.05 in log(S/N).

250 µm

10Extracted S/N

1

Sex

trac

ted/

Str

ue

6050.4


median

350 µm

10Extracted S/N

1

Sex

trac

ted/

Str

ue

6050.4


median

500 µm

10Extracted S/N

1

Sex

trac

ted/

Str

ue

6050.4


median

Figure 11. The ratio of flux densities for the matched sources in the simulated input (true) and extracted catalogues as a function of extracted signal to noisefor the three bands, using the gridded position simulations (including point sources only). Also shown are the median and 3σ clipped mean values, calculatedin bins of 0.05 in log(S/N). Note that confusion noise is not included in these simulations.

c© 2002 RAS, MNRAS 000, 1–13


250 µm; >5σ only

0.0 0.2 0.4 0.6 0.8 1.0S2nd brightest/Sbrightest

0

2.0•105

4.0•105

6.0•105

8.0•105

1.0•106

1.2•106

N

350 µm; >5σ only


0

1•105

2•105

3•105

4•105

5•105

6•105

N

500 µm; >5σ only


0

1•104

2•104

3•104

4•104

N

Figure 12. The PSF–weighted ratio of the brightest to second brightest input (true) source contributing to the extracted source, within the beam in each band,for > 5σ sources in the extracted catalogue.

250 µm

1 10Sall sources in beam/Strumatch

0

2•105

4•105

6•105

8•105

N

>5σ sources only>10σ sources only

> 1.5 = 10.5%

> 2.0 = 1.6%

> 2.5 = 0.2%

> 3.0 = 0.0%

350 µm


0

5.0•104

1.0•105

1.5•105

2.0•105

N


> 1.5 = 24.9%

> 2.0 = 5.7%

> 2.5 = 1.1%> 3.0 = 0.2%

500 µm


0

2000

4000

6000

8000

10000

N


> 1.5 = 56.5%

> 2.0 = 27.3%

> 2.5 = 10.4%

> 3.0 = 3.4%

Figure 13. The PSF–weighted, background–subtracted, ratio of the sum of simulated input (true) sources within a beam to the flux density of the matched truesource for > 5σ (solid line), and > 10σ (dashed line) sources in the extracted catalogue. The labels on the Figures give the percentage of sources with ratiosgreater than some particular value. The small proportion of sources where the ratio falls below 1 are due to the PSF–weighting.

S/N in the simulated input catalogue, which does not suffer fromthis effect, then they are in better agreement with the prediction.

Figure 8c also illustrates the improvement in positional er-rors that arises from selecting sources at 250µm only in MADX,instead of giving equal weight to all bands (flat–spectrum prior),as previously discussed in Section 3.1. Greater positional accuracysignificantly enhances the efficacy of the cross–identification to op-tical sources using the Likelihood Ratio method (Smith et al. 2011).This is why the better positions are deemed to outweigh the slightchance of missing red objects when using the 250µm prior.

4.2 Catalogue correction factors

Inspection of Figure 6 shows a clear discrepancy between the ex-tracted and simulated input integral counts at faint 500µm flux den-sities; this occurs due to a combination of two factors. The first,flux–boosting, is a preferential enhancement of faint source fluxdensities due to positive noise peaks, that arises due to the steep-ness of the faint end (i.e. S500µm ∼< 40 mJy; Clements et al. 2010)of the source counts. The second is a result of blending, where sev-eral simulated input sources (which may be too faint to be includedindividually) are detected as one source in the extracted catalogue.

These effects can be quantified by direct comparison of thesimulated input and extracted flux densities, shown in Figure 9, asa function of signal–to–noise in the extracted catalogue for boththe ESS and PSS versions. Flux correction factors are derived fromthe 3σ clipped mean of these data; these are given in Table 1. Ap-plying these factors to each extracted source gives a statistically‘flux–corrected’ catalogue. It should be noted however, that the dis-cussion of correction factors in this Section is restricted to sourcesdetected at a 5σ or greater level only.

An alternative approach to determining the catalogue correc-tion factors is to use a ‘noiseless’ catalogue, created by runningMADX on the simulated maps before the addition of noise, as thecomparison. As Figure 10 shows, this does not accurately repre-sent the level of flux–enhancement in the data, because, the noise-less catalogue is also affected by source blending. Additionally, atlow S/N the noiseless–input flux densities are generally brighterthan the extracted ones, suggesting that MADX underestimates thebackground subtraction in the absence of noise.

The relative contributions from the flux–boosting and sourceblending can be investigated with a new set of simulated, point–source only, maps, in which the sources are placed on a regu-lar spaced grid, with a 70′′ separation between points, to ensureno sources overlap. The source density is also lowered in thesemaps (imposed by excluding any source in the simulated input cat-alogue with a 250µm flux density fainter than 6.6 mJy), so thatsufficient unique positions can be generated. Inspecting the ratioof the extracted and simulated input fluxes – Figure 11 – suggeststhat the majority of the flux–enhancement seen in Figure 9 is dueto blended sources, rather than boosting due to noise. However,the PSF–weighted ratio of the brightest to second brightest simu-lated input source contributing to each source in the extracted cat-alogue (Figure 12) appears to contradict this; it shows that, even at500µm , blending with this second source would not increase theextracted flux density by the amount seen. The solution to this ap-parent contradiction becomes clear when the PSF–weighted ratioof the contribution from all the simulated input sources within abeam to the flux density of the simulated input match is consideredinstead (Figure 13). Here∼27% of 500µm> 5σ extracted sourceshave sufficient simulated input sources available to boost their fluxdensities by a factor of 2 or more when their contributions are com-

c© 2002 RAS, MNRAS 000, 1–13

10 E.E. Rigby

10 100Extracted S/N

0.01

0.10

1.00

Fra

ctio

nal f

lux

erro

r

250 µm350 µm500 µm


10 100Extracted S/N

0.01

0.10

1.00

Fra

ctio

nal f

lux

erro

r

250 µm350 µm500 µm


10 100True S/N

0.01

0.10

1.00

Fra

ctio

nal f

lux

erro

r

250 µm350 µm500 µm

(c) Point source simulations

Figure 14. The fractional flux density error for the corrected extracted cat-alogues, ignoring any sources that fall outside the 99.73rd percentiles. Thedotted lines indicate the expected behaviour.

bined, even though their individual effect is small. Figure 13 alsoshows that this confusion becomes negligible for > 10σ sources.This is in broad agreement with Chapin et al. (2011) who find thatthe sub–mm peaks they detect using a survey with larger beams,but of similar depth to H–ATLAS, generally consist of a blend ofseveral sources. Future versions of MADX will include a deblend-ing step which should reduce this effect. It should be noted that amean sky–background of 6.8 mJy, 5.8 mJy or 4.1 mJy at 250µm ,350µm and 500µm respectively (determined from the mean of thesimulated input catalogue), is subtracted before the histograms arecalculated, to account for the background–subtraction carried outas part of the source extraction process.

As a check on the success of the correction factors in Table 1,

they are applied to the full extracted catalogues and the fractionalflux density errors (after rejecting the points which lie outside the99.73rd percentile) are then calculated. As Figure 14 shows, thesereduce with increasing S/N, but, as with the positional errors dis-cussed previously, Eddington bias prevents this behaving exactly asexpected. Again, when plotted against the S/N from the simulatedinput catalogue (Figure 14c) the difference is reduced.

As well as the flux correction factors, we also need to com-pleteness of the detected catalogues, especially at faint 350 and500µm flux densities; this is clearly seen in Figures 15a and 16awhich compare the differential source counts for the extracted, sim-ulated input and flux–corrected catalogues. The lower counts aredue to the failure to detect some fraction of faint sources because ofrandom noise fluctuations in the simulated maps or source blend-ing. This incompleteness can be quantified by simply taking theratio of the flux–corrected to simulated input differential counts, togive a source–surface–density correction. Note that this is not ap-propriate for correcting the flux densities of individual sources, butrather it can be applied when making statistical analyses of the cat-alogue as a whole. This correction is shown in Figures 15b and 16b,and also given as an additional correction factor in Table 2. Figures15c and 16c demonstrate the success of the density correction whenapplied to the integral source counts.

There is one further factor that can affect the extracted cata-logue – contamination from spurious sources. The expected num-ber of > 5σ random noise peaks present in the 250µm map areais only ∼0.05, so this should be negligible in the SDP catalogue.Contamination from fainter sources which are boosted or blendedis accounted for in the flux correction factors.

It should be noted that an alternative approach to correct-ing the SDP H–ATLAS catalogue was adopted in Clements et al.(2010). In this case corrections were determined from the ratio ofextracted to simulated input integral source counts. This combinesthe effects of incompleteness and flux boosting, and is appropri-ate for recovering the correct source counts, but not for correctingindividual catalogue sources.

5 CONCLUDING REMARKS

This paper has presented the SDP catalogue for the first observa-tions of the H–ATLAS survey, along with a description of the sim-ulations created to determine the factors needed to correct it forthe combined effects of incompleteness, flux–boosting and sourceblending. The main results of this analysis are summarised below:

(i) The extracted flux densities of 350µm and 500µm sourcescan be enhanced over their simulated input values, by factors of upto ∼2. This predominantly affects sources with 5 < S/N < 15;

(ii) These enhancements are shown to be due to source blending,with ∼27% of > 5σ 500µm sources having sufficient simulatedinput sources available within a beam to create a boosting of ∼2;

(iii) A combination of flux density and source–surface–densitycorrections are necessary to correct the extracted source counts forthese factors.

It is anticipated that future development of the MADX soft-ware will incorporate subroutines to deal with both the effects ofmap pixelization and source blending in the processing stage.

MADX is not the only source extraction method being con-sidered for the H–ATLAS data, but time constraints mean that ithas been used for the SDP catalogue presented here. A comparison

c© 2002 RAS, MNRAS 000, 1–13


between different source extraction algorithms is currently ongo-ing; these include SUSSEXtractor developed by Savage & Oliver(2007), as well as the ‘matrix filter’ method of Herranz et al. (2009)and the ‘Mexican Hat wavelet’ method of Gonzalez-Nuevo et al.(2006) and Lopez-Caniego et al. (2006). The results of this com-parison will be used to improve future H–ATLAS catalogues.

This initial, uncorrected, catalogue will be available fromhttp://www.h-atlas.org, though it is expected that as thedata processing steps are refined it will undergo future updates.

ACKNOWLEDGMENTS

The Herschel-ATLAS is a project with Herschel, which isan ESA space observatory with science instruments providedby European-led Principal Investigator consortia and with im-portant participation from NASA. The H-ATLAS website ishttp://www.h-atlas.org/. U.S. participants in Herschel–ATLAS acknowledge support provided by NASA through a con-tract issued from JPL. The Italian group acknowledges partial fi-nancial support from ASI/INAF agreement n. I/009/10/0.

0.1 1.0S / Jy

100

101

102

103

104

105

dN/d

S

True catalogueExtracted catalogue

Flux-corrected catalogue

500 µm350 µm250 µm

(a) The differential source counts for the extracted, simulated input(true) and flux–corrected catalogues for the three bands. Note the dis-crepancy between the flux–corrected and simulated input cataloguesat faint flux densities.

0.1 1.0S / Jy

0.0

0.2

0.4

0.6

0.8

1.0

Den

sity

Cor

rect

ion

500 µm350 µm250 µm

(b) The surface density correction required at each band to correct forthe catalogue incompleteness, determined from the ratio of the flux–corrected to simulated input differential source counts. The solid dotsindicate the position of the average 5σ limit in each band.

0.1 1.0S / Jy

0.01

0.10

1.00

10.00

100.00

1000.00

N(>

S)

deg-2

500 µm350 µm250 µm

True catalogueFully corrected catalogue

(c) The integral source counts from the simulated input catalogueoverplotted with the flux and surface–density corrected catalogue todemonstrate the success of these correction factors at recovering thesimulated input values

Figure 15. Extended source simulations

c© 2002 RAS, MNRAS 000, 1–13

12 E.E. Rigby

0.1 1.0S / Jy

100

101

102

103

104

105

dN/d

S

True catalogueExtracted catalogue

Flux-corrected catalogue

500 µm350 µm250 µm

(a) The differential source counts for the extracted, simulated input(true) and flux–corrected catalogues for the three bands. Note the dis-crepancy between the flux–corrected and simulated input cataloguesat faint flux densities.

0.1 1.0S / Jy

0.0

0.2

0.4

0.6

0.8

1.0

Den

sity

Cor

rect

ion

500 µm350 µm250 µm

(b) The surface density correction required at each band to correct forthe catalogue incompleteness, determined from the ratio of the flux–corrected to simulated input differential source counts. The solid dotsindicate the position of the average 5σ limit in each band.

0.1 1.0S / Jy

0.01

0.10

1.00

10.00

100.00

1000.00

N(>

S)

deg-2

500 µm350 µm250 µm

True catalogueFully corrected catalogue

(c) The integral source counts from the simulated input catalogueoverplotted with the flux and surface–density corrected catalogue todemonstrate the success of these correction factors at recovering thesimulated input values

Figure 16. Point source simulations

c© 2002 RAS, MNRAS 000, 1–13


REFERENCES

Amblard, A., et al. 2010, A&A, 518, L9Baes, M., et al. 2010, A&A, 518, L39Berta, S., et al. 2010, A&A, 518, L30Bertin, E., & Arnouts, S. 1996, A&AS, 117, 393Chapin, E. L., et al. 2011, MNRAS, 411, 505Clements, D. L., et al. 2010, A&A, 518, L8Driver, S.P., et al., 2009, Astron. Geophys., 50, 5.12Dye, S., et al. 2010, A&A, 518, L10Eales, S., et al. 2010, PASP, 122, 499Gonzalez-Nuevo, J., Argueso, F., Lopez-Caniego, M., Toffolatti,

L., Sanz, J. L., Vielva, P., & Herranz, D. 2006, MNRAS, 369,1603

Gonzalez-Nuevo, J., et al. 2010, A&A, 518, L38Griffin, M. J., et al. 2010, A&A, 518, L3Herranz, D., Lopez-Caniego, M., Sanz, J. L., & Gonzalez-Nuevo,

J. 2009, MNRAS, 394, 510Ibar, E., 2010a, MNRAS, accepted, arXiv:1009.0262Ibar, E., 2010b, in prepIvison, R. J., et al. 2007, MNRAS, 380, 199Lopez-Caniego, M., Herranz, D., Gonzalez-Nuevo, J., Sanz, J. L.,

Barreiro, R. B., Vielva, P., Argueso, F., & Toffolatti, L. 2006,MNRAS, 370, 2047

Maddox, S. J., et al. 2010, A&A, 518, L11Maddox, S. et al., 2011, in prepNegrello, M., Magliocchetti, M., Moscardini, L., De Zotti, G.,

Granato, G. L., & Silva, L. 2004, MNRAS, 352, 493Negrello, M., Perrotta, F., Gonzalez-Nuevo, J., Silva, L., de Zotti,

G., Granato, G. L., Baccigalupi, C., & Danese, L. 2007, MN-RAS, 377, 1557

Nguyen, H. T., et al. 2010, A&A, 518, L5Ott, S. 2010, in ASP Conference Series, Astronomical Data Anal-

ysis Software and Systems XIX, Y. Mizumoto, K.-I. Morita, andM. Ohishi, eds., in press

Pascale, E., et al, 2010, MNRAS, submittedPilbratt, G. L., et al. 2010, A&A, 518, L1Poglitsch, A., et al. 2010 A&A, in press, arXiv:1005.1487Poglitsch, A., et al. 2010, A&A, 518, L2Savage, R. S., & Oliver, S. 2007, ApJ, 661, 1339Serjeant, S., et al. 2003, MNRAS, 344, 887Serjeant, S., et al. 2010, A&A, 518, L7Schlegel, D. J., Finkbeiner, D. P., & Davis, M. 1998, ApJ, 500,

525Smith, D. J. B., et al. 2011, MNRAS submitted, arXiv:1007.5260Thompson, M. A., et al. 2010, A&A, 518, L134Turin, G. 1960, IRE Transactions on Information Theory, 6, 311Viero, M. P., et al. 2009, ApJ, 707, 1766Wang, L., & Rowan-Robinson, M. 2009, MNRAS, 398, 109Zhong, G. H., Liang, Y. C., Liu, F. S., Hammer, F., Hu, J. Y., Chen,

X. Y., Deng, L. C., & Zhang, B. 2008, MNRAS, 391, 986

This paper has been typeset from a TEX/ LATEX file prepared by theauthor.

c© 2002 RAS, MNRAS 000, 1–13

http://arxiv.org/abs/1009.0262



14 E.E. Rigby

ESS PSSCatalogue S/N FC250µm FC350µm FC500µm FC250µm FC350µm FC500µm

5.30 1.06 1.12 1.45 1.06 1.12 1.455.94 1.06 1.18 1.51 1.06 1.18 1.506.67 1.06 1.21 1.51 1.06 1.21 1.507.48 1.06 1.23 1.47 1.06 1.23 1.458.39 1.06 1.25 1.35 1.06 1.25 1.329.42 1.06 1.27 1.14 1.06 1.26 1.1110.57 1.06 1.27 1.01 1.06 1.25 1.0211.86 1.05 1.23 1.00 1.05 1.20 1.0013.30 1.04 1.15 0.99 1.04 1.10 0.9914.93 1.02 1.04 0.98 1.02 1.01 0.9916.75 1.00 1.01 0.98 1.00 0.99 0.9818.79 0.98 1.00 0.98 0.98 0.99 0.9921.08 0.97 0.99 0.97 0.98 0.98 0.9823.66 0.96 0.99 0.97 0.97 0.98 0.9826.54 0.96 0.99 0.96 0.97 0.98 0.9829.78 0.96 0.98 0.96 0.97 0.97 0.9833.42 0.96 0.98 0.98 0.97 0.98 0.9837.49 0.95 0.98 0.97 0.97 0.97 0.9942.07 0.96 0.98 0.96 0.97 0.97 1.0047.20 0.95 0.97 0.99 0.97 0.97 0.9852.96 0.95 0.97 0.96 0.97 0.97 0.9959.43 0.96 0.94 0.95 0.97 0.97 0.9766.68 0.95 0.93 – 0.96 0.97 –74.81 0.95 0.92 – 0.97 0.97 –83.94 0.97 0.93 – 0.97 0.97 –94.18 0.97 0.94 – 0.96 0.97 –

Table 1. The flux density correction factors (FC) at each SPIRE wavelength, as a function of S/N in the extracted catalogue, determined from the ratio of fluxdensities in the matched extracted and simulated input catalogues. To apply the correction at some catalogue flux density, fcat: fcorr = fcat/FC, thoughnote that the density correction given in Table 2 should also be applied as well.

c© 2002 RAS, MNRAS 000, 1–13


Corrected flux density ESS PSS(Jy) SC250µm SC350µm SC500µm SC250µm SC350µm SC500µm

0.0320 0.31 – – 0.40 – –0.0327 0.75 – 0.11 0.79 – 0.110.0335 0.84 – 0.41 0.85 – 0.390.0343 0.84 – 0.49 0.85 – 0.490.0351 0.83 – 0.48 0.85 – 0.480.0359 0.85 0.01 0.46 0.86 0.01 0.460.0367 0.84 0.68 0.39 0.86 0.71 0.410.0376 0.85 1.36 0.31 0.87 1.36 0.330.0385 0.83 1.36 0.26 0.86 1.36 0.270.0394 0.83 1.34 0.25 0.86 1.34 0.260.0403 0.85 1.07 0.25 0.88 1.11 0.270.0427 0.85 0.84 0.21 0.88 0.86 0.240.0490 0.84 0.68 0.19 0.87 0.71 0.220.0562 0.84 0.57 0.22 0.88 0.60 0.260.0646 0.83 0.45 0.29 0.85 0.49 0.290.0741 0.82 0.44 0.41 0.83 0.43 0.380.0851 0.82 0.49 0.49 0.84 0.43 0.590.0977 0.84 0.58 0.61 0.87 0.55 0.790.1122 0.89 0.74 0.74 0.92 0.70 0.930.1288 0.95 0.84 0.78 1.00 0.91 0.930.1479 0.93 0.87 0.72 1.00 1.05 0.960.1698 0.89 0.81 0.67 0.99 0.99 0.990.1950 0.93 0.77 0.72 1.01 0.99 0.990.2239 0.87 0.82 0.87 1.00 1.00 1.000.2570 0.90 0.83 0.77 1.00 1.00 1.000.2951 0.95 0.92 0.92 1.00 1.00 1.000.3388 0.96 0.89 0.73 1.00 1.00 1.000.3890 0.94 0.92 1.00 1.00 1.00 1.000.4467 0.98 1.00 1.00 1.00 1.00 1.000.5129 0.93 1.00 1.00 1.00 1.00 1.000.5888 1.02 1.00 1.00 1.00 1.00 1.000.6761 1.00 1.00 1.00 1.00 1.00 1.000.7762 1.00 1.00 1.00 1.00 1.00 1.000.8913 1.00 1.00 1.00 1.00 1.00 1.001.0233 1.00 1.00 1.00 1.00 1.00 1.00

Table 2. The surface density correction (SC) at each SPIRE wavelength as a function of corrected flux density, determined from the ratio of the flux–correctedto simulated input differential counts. To apply the correction at some corrected flux density, fcorr: fcorr final = fcorr/SC. The corrected flux densitiesgiven are the central bin values.

c© 2002 RAS, MNRAS 000, 1–13

Herschel–ATLAS: First data release of the Science ... · Herschel–ATLAS: First data release of...

Documents

Transcript of Herschel–ATLAS: First data release of the Science ... · Herschel–ATLAS: First data release of...